News
Top 10: Future AI Leaders
Soujanya Poria is an assistant professor of Computer Science at the Singapore University of Technology and Design (SUTD). He has co-authored more than 150 papers, published in top-tier conferences and journals.
For discovering three security flaws in Qualcomm chips, SUTD researcher received over $10,000
Students from the Singapore University of Technology and Design discovered three security vulnerabilities in Qualcomm’s 5G chips, which may disable the 5G network services of about 60% of the world’s mobile phones.
‘5Ghoul’ Vulnerabilities Haunt Qualcomm, MediaTek 5G Modems
Collectively tagged as 5Ghoul, the 14 security defects can be exploited to drop and freeze 5G connections on smartphones and routers, and to conduct downgrading attacks, according to the research team. The majority of the flaws affect 5G modems from Qualcomm and MediaTek.
Security flaw discovered in Qualcomm chips
Channel NewsAsia speaks to SUTD researcher Dr Matheus Garbelini who shared about the discovery process. This discovery was made by Dr Garbelini and his team, which includes Assistant Prof Sudipta Chattopadhyay, SUTD PhD student, Shang Zewen, and two researchers from A*STAR.
New local AI music generation model produces “first draft” for composers in one minute
The first “controllable text prompt” AI music generation model has been developed locally. The software called Mustango was developed by a six-person research team led by Soujanya Poria and Dorien Herremans, two assistant professors at the Singapore University of Technology and Design.
A novel framework for smooth avatar-user synchronization in the metaverse
Assistant Professor Xiong Zehui and his collaborators from NTU and GDUT proposed a novel framework to optimize resource allocation in metaverse service providers, with the overall aim of ensuring a smooth and immersive experience for all users.
Meet TinyLlama: The 550MB AI Model Trained on 3 Trillion Tokens
The TinyLlama project, by researchers from the StatNLP Research Group at SUTD, is trying to pre-train a 1.1 billion Llama model on a whopping three trillion tokens. This model takes up only 550MB of RAM.
Keeping the backdoor secure in your robust machine learning model
In the study ‘Towards backdoor attacks and defense in robust machine learning models’ Assistant Prof Sudipta Chattopadhyay and fellow SUTD researchers studied how to inject and defend against backdoor attacks for robust models in a certain ML component called image classifiers.
Six SUTD Faculty, Alumni and Student in Forbes 30 Under 30 Asia Class of 2023
Six SUTD faculty, alumni and student join the 300 outstanding young trailblazers of Forbes 30 Under 30 Asia Class of 2023.
TikTok bans: will Asia follow West’s lead as US turns the screws on China-linked app?
Assistant Professor Roy Lee (Information Systems Technology and Design) shares his expert opinion on the topic.
